Coursera Spring Sale
40% Off Coursera Plus Annual!
Grab it
Explore a groundbreaking conference talk that presents innovative hybrid physics and artificial intelligence modeling approaches for climate simulation. Learn how researchers at the California Institute of Technology are addressing one of climate science's most persistent challenges - the uncertainty in cloud feedback estimates - through the development of the CliMA atmosphere-land climate model. Discover how machine learning components are integrated with traditional process models to simulate turbulence, clouds, and convection more accurately. Understand the novel calibration methodology that uses global Earth observations to constrain model parameters, moving beyond post-facto filtering approaches to directly calibrate climate models with observational data. Examine how this approach produces a posterior distribution over uncertain model parameters, enabling the generation of observationally-calibrated model ensembles that more faithfully reproduce present climate conditions and variability. Gain insights into how this methodology integrates observational uncertainty directly into the model framework, providing robust climate change projections with quantified uncertainty bounds for cloud feedbacks and other critical climate processes.
